R help - Aug 2008 - missing TRUE/FALSE error in conditional construct

Hi everyone,

I posted something similar to this in reply to another post, but there seems
to be a problem getting it onto the board, so I'm starting a new post.

I am trying to use conditional formatting to select non-zero and non-NaN
values from a matrix and pass them into another matrix. The problem is that
I keep encountering an error message indicating the ":missing value where
TRUE/FALSE needed "

My code is as follows:

##Code Start 
mat_zeroless<-matrix(NA,5000,2000) #generating holding matrix 

##Assume source matrix containing combination of values, NaNs and zeros##
for (j in 1:5000) 
        { 
for (k in 1:2000) 
{ 
if(mat[j,k]!=0 & !is.NaN(mat[j,k])) {mat_zeroless[j,k]<-mat[j,k]} 
} 
        } 
##Code End 

Error in if (mat[j,k] !=0 & !is.NaN(mat[j,k])) { :missing value where
TRUE/FALSE needed 

I'm not sure how to resolve this.

rcoder 
-- 
View this message in context:
http://www.nabble.com/missing-TRUE-FALSE-error-in-conditional-construct-tp18972244p18972244.html
Sent from the R help mailing list archive at Nabble.com.

> -----Original Message-----
> From: r-help-bounces at r-project.org
[mailto:r-help-bounces at r-project.org]
> On Behalf Of rcoder
> Sent: Wednesday, August 13, 2008 3:17 PM
> To: r-help at r-project.org
> Subject: [R] missing TRUE/FALSE error in conditional construct
> 
> 
> Hi everyone,
> 
> I posted something similar to this in reply to another post, but there
> seems
> to be a problem getting it onto the board, so I'm starting a new post.
> 
> I am trying to use conditional formatting to select non-zero and
non-NaN
> values from a matrix and pass them into another matrix. The problem is
> that
> I keep encountering an error message indicating the ":missing value
where
> TRUE/FALSE needed "
> 
> My code is as follows:
> 
> ##Code Start
> mat_zeroless<-matrix(NA,5000,2000) #generating holding matrix
> 
> ##Assume source matrix containing combination of values, NaNs and
zeros##
> for (j in 1:5000)
>         {
> for (k in 1:2000)
> {
> if(mat[j,k]!=0 & !is.NaN(mat[j,k])) {mat_zeroless[j,k]<-mat[j,k]}
> }
>         }
> ##Code End
> 
> Error in if (mat[j,k] !=0 & !is.NaN(mat[j,k])) { :missing value where
> TRUE/FALSE needed
> 
> I'm not sure how to resolve this.
This seems to do what you appear to need,
no loops required (always best whenever possible):
> set.seed(123)
> mat <- matrix(sample(1:10), nrow = 5)
> mat
     [,1] [,2]
[1,]    3    1
[2,]    8   10
[3,]    4    9
[4,]    7    2
[5,]    6    5
> is.na(mat) <- c(2, 3)
> mat
     [,1] [,2]
[1,]    3    1
[2,]   NA   10
[3,]   NA    9
[4,]    7    2
[5,]    6    5
> mat[5,] <- 0
> mat
     [,1] [,2]
[1,]    3    1
[2,]   NA   10
[3,]   NA    9
[4,]    7    2
[5,]    0    0
> matno0 <- matrix(NA, nrow = nrow(mat), ncol = ncol(mat))
> matno0
     [,1] [,2]
[1,]   NA   NA
[2,]   NA   NA
[3,]   NA   NA
[4,]   NA   NA
[5,]   NA   NA
> mat[!is.na(mat) & !(mat == 0)]
[1]  3  7  1 10  9  2
> matno0[!is.na(mat) & !(mat == 0)] <- mat[!is.na(mat) & !(mat ==
0)]
> matno0
     [,1] [,2]
[1,]    3    1
[2,]   NA   10
[3,]   NA    9
[4,]    7    2
[5,]   NA   NA
>
Other issues:

The error messages you are seeing generally
are happening because you are generating
non-scalar TRUE / FALSE outcomes or
NA outcomes in your if() clause.

An if() clause should generate scalar
TRUE or FALSE only.

To this end, the S language provides
the operators '&&' (scalar AND)
and '||' (scalar OR) for use in
logical scalar clauses.  Further,
if the first argument to '&&' tests
FALSE, the second is not even evaluated,
so you could have done

if ( !is.na(mat[j, k]) && mat[j, k] != 0 ) { blah }

In this case, once the NA is found, the if() clause
evaluation is over and you don't have to worry about
the NA that will result from the second argument
mat[j, k] != 0

 
> matno0 <- matrix(NA, nrow = nrow(mat), ncol = ncol(mat))
> matno0
     [,1] [,2]
[1,]   NA   NA
[2,]   NA   NA
[3,]   NA   NA
[4,]   NA   NA
[5,]   NA   NA
> for ( j in 1:5 ) {
+  for ( k in 1:2 ) {
+   if ( !is.na(mat[j, k]) && mat[j, k] != 0 ) { matno0[j, k] <-
mat[j,
k] }
+  }
+ }
> 
> matno0
     [,1] [,2]
[1,]    3    1
[2,]   NA   10
[3,]   NA    9
[4,]    7    2
[5,]   NA   NA
> 
HTH

Steve McKinney
> 
> rcoder
> --
> View this message in context:
http://www.nabble.com/missing-TRUE-FALSE-
> error-in-conditional-construct-tp18972244p18972244.html
> Sent from the R help mailing list archive at Nabble.com.
> 
> ______________________________________________
> R-help at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide http://www.R-project.org/posting-
> guide.html
> and provide commented, minimal, self-contained, reproducible code.

The problem is that if x is either NA or NaN then x != 0 is NA (and not FALSE or
TRUE) and the function is.nan tests for a NaN but not for NA, i.e. is.nan(NA)
returns FALSE.

You can do something like:

mat_zeroless[!is.na(mat) & mat != 0] <- mat[!is.na(mat) & mat != 0]


--- On Thu, 14/8/08, rcoder <mpdotbook at gmail.com> wrote:
> From: rcoder <mpdotbook at gmail.com>
> Subject: [R]  missing TRUE/FALSE error in conditional construct
> To: r-help at r-project.org
> Received: Thursday, 14 August, 2008, 8:16 AM
> Hi everyone,
> 
> I posted something similar to this in reply to another
> post, but there seems
> to be a problem getting it onto the board, so I'm
> starting a new post.
> 
> I am trying to use conditional formatting to select
> non-zero and non-NaN
> values from a matrix and pass them into another matrix. The
> problem is that
> I keep encountering an error message indicating the
> ":missing value where
> TRUE/FALSE needed "
> 
> My code is as follows:
> 
> ##Code Start 
> mat_zeroless<-matrix(NA,5000,2000) #generating holding
> matrix 
> 
> ##Assume source matrix containing combination of values,
> NaNs and zeros##
> for (j in 1:5000) 
>         { 
> for (k in 1:2000) 
> { 
> if(mat[j,k]!=0 & !is.NaN(mat[j,k]))
> {mat_zeroless[j,k]<-mat[j,k]} 
> } 
>         } 
> ##Code End 
> 
> Error in if (mat[j,k] !=0 & !is.NaN(mat[j,k])) {
> :missing value where
> TRUE/FALSE needed 
> 
> I'm not sure how to resolve this.
> 
> rcoder 
> -- 
> View this message in context:
>
http://www.nabble.com/missing-TRUE-FALSE-error-in-conditional-construct-tp18972244p18972244.html
> Sent from the R help mailing list archive at Nabble.com.
> 
> ______________________________________________
> R-help at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide
> http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained,
> reproducible code.